mongoid-sphinx 0.0.3 → 0.0.4

data/lib/mongoid_sphinx.rb

@@ -2,7 +2,5 @@ require "mongoid"
 require "riddle"
 
 require 'mongoid_sphinx/configuration'
-require 'mongoid_sphinx/indexer'
-require 'mongoid_sphinx/multi_attribute'
 require 'mongoid_sphinx/mongoid/identity'
 require 'mongoid_sphinx/mongoid/sphinx'

data/lib/mongoid_sphinx/mongoid/sphinx.rb

@@ -5,18 +5,72 @@ module Mongoid
   module Sphinx
     extend ActiveSupport::Concern
     included do
+      SPHINX_TYPE_MAPPING = {
+        'Date' => 'timestamp',
+        'DateTime' => 'timestamp',
+        'Time' => 'timestamp',
+        'Float' => 'float',
+        'Integer' => 'int',
+        'BigDecimal' => 'float',
+        'Boolean' => 'bool'
+      }
+      
       cattr_accessor :search_fields
+      cattr_accessor :search_attributes
     end
     
     module ClassMethods
-      def search_index(*fields)
-        self.search_fields = fields
+      def search_index(options={})
+        self.search_fields = options[:fields]
+        self.search_attributes = {}
+        options[:attributes].each do |attrib|
+          self.search_attributes[attrib] = SPHINX_TYPE_MAPPING[self.fields[attrib.to_s].type.to_s] || 'str2ordinal'
+        end
+      end
+      
+      def sphinx_stream
+        STDOUT.sync = true # Make sure we really stream..
+        
+        puts '<?xml version="1.0" encoding="utf-8"?>'
+        puts '<sphinx:docset>'
+        
+        # Schema
+        puts '<sphinx:schema>'
+        puts '<sphinx:field name="classname"/>'
+        self.search_fields.each do |key, value|
+          puts "<sphinx:field name=\"#{key}\"/>"
+        end
+        self.search_attributes.each do |key, value|
+          puts "<sphinx:attr name=\"#{key}\" type=\"#{value}\"/>"
+        end
+        puts '</sphinx:schema>'
+        
+        self.all.each do |document_hash|
+          sphinx_compatible_id = document_hash['_id'].to_s.to_i - 100000000000000000000000
+          
+          puts "<sphinx:document id=\"#{sphinx_compatible_id}\">"
+          
+          puts "<classname>#{self.to_s}</classname>"
+          self.search_fields.each do |key|
+            puts "<#{key}><![CDATA[[#{document_hash[key.to_s]}]]></#{key}>"
+          end
+          self.search_attributes.each do |key, value|
+            value = case value
+              when 'bool' : document_hash[key.to_s] ? 1 : 0
+              when 'timestamp' : document_hash[key.to_s].to_i
+              else document_hash[key.to_s].to_s
+            end 
+            puts "<#{key}>#{value}</#{key}>"
+          end
+          
+          puts '</sphinx:document>'
+        end
+        
+        puts '</sphinx:docset>'
       end
       
       def search(query, options = {})
         client = MongoidSphinx::Configuration.instance.client
-                 
-        query = query + " @classname #{self.to_s}"
         
         client.match_mode = options[:match_mode] || :extended
         client.limit = options[:limit] if options.key?(:limit)
@@ -27,17 +81,27 @@ module Mongoid
           client.sort_by = options[:sort_by]
         end
         
-        result = client.query(query)
+        if options.key?(:with)
+          options[:with].each do |key, value|
+            client.filters << Riddle::Client::Filter.new(key.to_s, value.is_a?(Range) ? value : value.to_a, false)
+          end
+        end
+        
+        if options.key?(:without)
+          options[:without].each do |key, value|
+            client.filters << Riddle::Client::Filter.new(key.to_s, value.is_a?(Range) ? value : value.to_a, true)
+          end
+        end
+        
+        result = client.query("#{query} @classname #{self.to_s}")
         
         if result and result[:status] == 0 and (matches = result[:matches])
-          classname = nil
           ids = matches.collect do |row|
-            classname = MongoidSphinx::MultiAttribute.decode(row[:attributes]['csphinx-class'])
             (100000000000000000000000 + row[:doc]).to_s rescue nil
           end.compact
           
           return ids if options[:raw] or ids.empty?
-          return Object.const_get(classname).find(ids)
+          return self.find(ids)
         else
           return []
         end

data/lib/mongoid_sphinx/version.rb

@@ -1,3 +1,3 @@
 module MongoidSphinx #:nodoc
-  VERSION = "0.0.3"
+  VERSION = "0.0.4"
 end

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: mongoid-sphinx
 version: !ruby/object:Gem::Version 
-  hash: 25
+  hash: 23
   prerelease: false
   segments: 
   - 0
   - 0
-  - 3
-  version: 0.0.3
+  - 4
+  version: 0.0.4
 platform: ruby
 authors: 
 - Matt Hodgson
@@ -63,10 +63,8 @@ extra_rdoc_files: []
 
 files: 
 - lib/mongoid_sphinx/configuration.rb
-- lib/mongoid_sphinx/indexer.rb
 - lib/mongoid_sphinx/mongoid/identity.rb
 - lib/mongoid_sphinx/mongoid/sphinx.rb
-- lib/mongoid_sphinx/multi_attribute.rb
 - lib/mongoid_sphinx/version.rb
 - lib/mongoid_sphinx.rb
 - README.markdown

data/lib/mongoid_sphinx/indexer.rb

@@ -1,123 +0,0 @@
-# MongoidSphinx, a full text indexing extension for MongoDB using
-# Sphinx.
-#
-# This file contains the MongoidSphinx::Indexer::XMLDocset and
-# MongoidSphinx::Indexer::XMLDoc classes.
-
-module MongoidSphinx #:nodoc:
-
-  # Module Indexer contains classes for creating XML input documents for the
-  # indexer. Each Sphinx index consists of a single "sphinx:docset" with any
-  # number of "sphinx:document" tags.
-  #
-  # The XML source can be generated from an array of CouchRest objects or from
-  # an array of Hashes containing at least fields "classname" and "_id"
-  # as returned by MongoDB view "MongoSphinxIndex/couchrests_by_timestamp".
-  #
-  # Sample:
-  #
-  #   rows = [{ 'name' => 'John', 'phone' => '199 43828',
-  #             'classname' => 'Address', '_id' => 'Address-234164'
-  #           },
-  #           { 'name' => 'Sue', 'mobile' => '828 19439',
-  #             'classname' => 'Address', '_id' => 'Address-422433'
-  #          }
-  #   ]
-  #   puts MongoSphinx::Indexer::XMLDocset.new(rows).to_s
-  #
-  #   <?xml version="1.0" encoding="utf-8"?>
-  #   <sphinx:docset>
-  #     <sphinx:schema>
-  #       <sphinx:attr name="csphinx-class" type="multi"/>
-  #       <sphinx:field name="classname"/>
-  #       <sphinx:field name="name"/>
-  #       <sphinx:field name="phone"/>
-  #       <sphinx:field name="mobile"/>
-  #       <sphinx:field name="created_at"/>
-  #     </sphinx:schema>
-  #     <sphinx:document id="234164">
-  #       <csphinx-class>336,623,883,1140</csphinx-class>
-  #       <classname>Address</classname>
-  #       <name><![CDATA[[John]]></name>
-  #       <phone><![CDATA[[199 422433]]></phone>
-  #       <mobile><![CDATA[[]]></mobile>
-  #       <created_at><![CDATA[[]]></created_at>
-  #     </sphinx:document>
-  #     <sphinx:document id="423423">
-  #       <csphinx-class>336,623,883,1140</csphinx-class>
-  #       <classname>Address</classname>
-  #       <name><![CDATA[[Sue]]></name>
-  #       <phone><![CDATA[[]]></phone>
-  #       <mobile><![CDATA[[828 19439]]></mobile>
-  #       <created_at><![CDATA[[]]></created_at>
-  #     </sphinx:document>
-  #   </sphinx:docset>"
-
-  module Indexer
-
-    # Class XMLDocset wraps the XML representation of a document to index. It
-    # contains a complete "sphinx:docset" including its schema definition.
-    
-    class XMLDocset
-
-      # Streams xml of all objects in a klass to the stdout. This makes sure you can process large collections.
-      #
-      # Options:
-      #  attributes (required) - The attributes that are put in the sphinx xml.
-      #
-      # Example:
-      #  MongoSphinx::Indexer::XMLDocset.stream(Document, :attributes => %w(title content))
-      # This will create an XML stream to stdout. 
-      #
-      # Configure in your sphinx.conf like
-      #  xmlpipe_command = ./script/runner "MongoSphinx::Indexer::XMLDocset.stream(Document, :attributes => %w(title content))"
-      #
-      def self.stream(klass)
-        STDOUT.sync = true # Make sure we really stream..
-
-        puts '<?xml version="1.0" encoding="utf-8"?>'
-        puts '<sphinx:docset>'
-
-        # Schema
-        puts '<sphinx:schema>'
-        klass.search_fields.each do |key, value|
-          puts "<sphinx:field name=\"#{key}\"/>"
-        end
-        # FIXME: What is this attribute?
-        puts '<sphinx:field name="classname"/>'
-        puts '<sphinx:attr name="csphinx-class" type="multi"/>'
-        puts '</sphinx:schema>'
-        
-        collection = Mongoid.database.collection(klass.collection.name)
-        collection.find.each do |document_hash|
-          XMLDoc.stream_for_hash(document_hash, klass)
-        end
-
-        puts '</sphinx:docset>'
-      end
-      
-    end
-
-    class XMLDoc
-
-      def self.stream_for_hash(hash, klass)
-        sphinx_compatible_id = hash['_id'].to_s.to_i - 100000000000000000000000
-        
-        puts "<sphinx:document id=\"#{sphinx_compatible_id}\">"
-        # FIXME: Should we include this?
-        puts '<csphinx-class>'
-        puts MongoidSphinx::MultiAttribute.encode(klass.to_s)
-        puts '</csphinx-class>'
-        puts "<classname>#{klass.to_s}</classname>"
-        
-        klass.search_fields.each do |key|
-          value = hash[key.to_s]
-          puts "<#{key}><![CDATA[[#{value}]]></#{key}>"
-        end
-
-        puts '</sphinx:document>'
-      end
-      
-    end
-  end
-end

data/lib/mongoid_sphinx/multi_attribute.rb

@@ -1,57 +0,0 @@
-# MongoSphinx, a full text indexing extension for MongoDB using
-# Sphinx.
-#
-# This file contains the MongoSphinx::MultiAttribute class.
-
-# Namespace module for the MongoSphinx gem.
-
-module MongoidSphinx #:nodoc:
-
-  # Module MultiAttribute implements helpers to translate back and
-  # forth between Ruby Strings and an array of integers suitable for Sphinx
-  # attributes of type "multi".
-  #
-  # Background: Getting an ID as result for a query is OK, but for example to
-  # allow cast safety, we need an aditional attribute. Sphinx supports
-  # attributes which are returned together with the ID, but they behave a
-  # little different than expected: Instead we can use arrays of integers with
-  # ASCII character codes. These values are returned in ascending (!) order of
-  # value (yes, sounds funny but is reasonable from an internal view to
-  # Sphinx). So we mask each byte with 0x0100++ to keep the order...
-  #
-  # Sample:
-  #
-  #   MongoSphinx::MultiAttribute.encode('Hello')
-  #   => "328,613,876,1132,1391"
-  #   MongoSphinx::MultiAttribute.decode('328,613,876,1132,1391')
-  #   => "Hello"
-
-  module MultiAttribute
-
-    # Returns an numeric representation of a Ruby String suitable for "multi"
-    # attributes of Sphinx.
-    #
-    # Parameters:
-    #
-    # [str] String to translate
-
-    def self.encode(str)
-      offset = 0
-      return str.bytes.collect { |c| (offset+= 0x0100) + c }.join(',')
-    end
-
-    # Returns the original MongoDB ID created from a Sphinx ID. Only works if
-    # the ID was created from a MongoDB ID before!
-    #
-    # Parameters:
-    #
-    # [multi] Sphinx "multi" attribute to translate back
-
-    def self.decode(multi)
-      offset = 0
-      multi = multi.split(',') if not multi.kind_of? Array
-
-      return multi.collect {|x| (x.to_i-(offset+=0x0100)).chr}.to_s
-    end
-  end
-end